Enzymatic Deacylation of Mono-and Dibutyryl Derivatives of Cyclic Adenosine 3’, S-Monophosphate by Extracts of Rat Tissues*

SUMMARY Cell-free extracts of several rat tissues contained NG-butyryl amidohydrolase and OS’-butyryl esterase activities for the deacylation of [3H]butyryl-labeled NC, 02’-dibutyryl cyclic AMP, N6-monobutyryl cyclic AMP, and 02’-monobutyryl cyclic AMP. The highest rates of deacylation were in the soluble cyto-plasm of all tissues; cyclic AMP phosphodiesterase activities were also highest in the same cell compartment. Since rates of cyclic AMP phosphodiesterase and 5’-AMP 5’-nucleotidase activities were from one to two orders of magnitude greater than those of the deacylases, deacylation is clearly the rate-limiting process in the catabolism of butyryl derivatives of cyclic AMP to adenosine. relatively of deacylation, they are sufficiently high to produce, even when operating minimally, amounts of cyclic AMP far greater than could be produced endogenously, even under hormonal of intracellular cyclic AMP of biologically active monobutyryl

effectiveness of these derivatives, and little is known about their catabolism.
Henion et al. (6) proposed two csplanations for their mechanism of action: (a) the derivatives may penetrate biological membranes more readily than cXhIP, and (b) the derivatives may be rnore refractory to the xtion of the catabolic enzyme, cXMP I)hosphodiest~rasr.
In the only direct study of mechanism (a) reported, Kaukel and Hilz (7), in a study published following completion of our work, found that estraccllular [3H]c.U11' was rapidly and completely degraded by enzymes located on the surface of HeLa cells and in the tissue culture medium to a mixture of 3H-labeled purine metabolites; these compounds readily entered the cells and, to a small degree, were reconverted to [%]L~MP. In contrast, ring-labeled N6,02'-[3H]dibutyryl ciiM1' resisted estracellula,r degradation, aas rapidly taken up by the cells, and was mostly converted illtracellularly to N6-[3H]rnonobutyr3-] cLIMP. With regard to mechanism (b) above, resistance of butyryl tlerivatives of cAMI' to catabolism by cXVl1' phosphodiesteranc has been well documentcd in seyeral laboratories (6,(8)(9)(10).
With regard to the mcchanisrn of action and metabolism of these acyl derivatives of c,l?\IP, many quest,ions remain manswered, among which arc the following.
In this paper, we show dirtctly that crllfree extracts of rat adipose, brain, heart, kidney, and li\-car ~011. tain N6-butyryl amidohydrolases and 02'-butyryl esterases which deacylated B&CAMP, N6-Bt-CAMP, and 02'-Bt-CAMP to CAMP, but which did so at rates very much lower than those of the CAMP phosphodiesterase in the same tissue compartments. Furthcrmorc, these rates of deacylation were of a magnitude sufficient to produce biologically effective amounts of CAMP. Some of these observations have been reported in preliminary form (5).
[3H]Butyric acid was removed from the solution by three extractions with three volumes each of diethyl ether, following which the pH of the aqueous phase was quickly adjusted to 6.5. The hydrolysate was lyophilized, and the nucleotides in the residue were separated and isolated by preparative paper chromatography as we described before (12) (14). Cyclic AMP (BMC lot 6499415) and freshly prepared 4-morpholino-N , N'dicyclohexyl carboxamidine (15), each 0.3 mmole, were refluxed in 3 ml of dry pyridine until complete dissolution occurred (15 to 30 min) ; the reaction mixture was then returned to 25". To this solution was added 0.25 mmole (12.5 mCi) of [2,3-3H]butyric anhydride (custom synthesis by Schwarz-Mann, lot XR-2051) and 8.86 mmoles of unlabeled butyric anhydride, and the mixture was stirred under anhydrous conditions for 17 hours; the reaction was ended by the addition of 1.5 ml of water at icebath temperature. The esterification was followed by periodically transferring I-$ aliquots to 4 ~1 of water, and, after 4 hours of hydrolysis, chromatographing this solution on thin layers of cellulose (Machery-Nagel MN 300, A254) in ethanol-l M ammonium acetate (7:3). The only radioactive ultravioletabsorbing spots were those with mobilities (RF 0.93 and 0.73, respectively) corresponding to authentic (BMC) B&CAMP and Bt-CAMP (N6-Bt-CAMP and 02'-Bt-CAMP have identical mobilities in this system; Fig. 2 and Reference 14). The reaction mixture was dried at 40" under reduced pressure (rotary evaporator), and the residue was subjected to preparative paper chromatography (12). 02'-[3H]Bt-cAMP was identified by the following: itsmobility (RF 0.73) in the thin layer chromatographic system compared to that of authentic N6-Bt-CAMP, its ultraviolet absorption spectrum at pH 7 ( Fig. 1, Curve 2) which shows a maximum at 257 nm like that of cilhIP but in contrast to the absorption maximum at 274 nm observed with N6-acyl derivatives ( Fig. 1, Curses S to 5), its extinction quotients at pH 7 (250/260, 0.85; 280/260, 0.16; 290/260, 0.01 (no literature values available, but compare with the corresponding quotients for N6-Bt-CAMP of 0.73, 1.15, and 0.30, respectively, according to BMC data)), and its quantitative conversion to CAMP upon brief exposure to alkali ( Fig. 2; cf. Reference 11). The specific radioactivity of a typical preparation was 1570 dpm per nmole.
Tissue Preparations-Tissues from male, Sprague-Dawley rats (140 to 200 g, fed ad Zibitum) were homogenized in 0.25 M sucrose-40 mM 'Iris-HCl (pH 7.4) in a loose-fitting, all glass, motor-driven (725 rpm) Potter-Elvehjem instrument (Kontes) for 10 to 15 excursions. Homogenates, except adipose tissue, were centrifuged briefly at 100 X g to remove debris, then at 104,500 x g for 30 min (4") ; particulates were washed three times by homogenization in buffer, and the washings (104,500 x g supernatant fluids) were pooled with the cytosols. In the case of adipose tissue, the homogenate was centrifuged for 10 min at 2000 x g at 4"; the infranatant fluid (under the solidified fat cake at the surface) was then subjected to fractionation as for the other tissues.
Protein was determined by a modified biuret procedure (16).
Enzyme Assays c&UP Phos$odiesterase-The assay, a modification (17) of that of Butcher and Sutherland (18), measures both high and low K, enzymes. Incubation mixtures contained, in a total volume of 0.5 ml of Tris-HCl buffer (pH 7.5), 0.88 pmole of MgSO+ excess (1 to 2 pmoles) CAMP, and various amounts of particulate or cytosol protein. Zero order kinetics were obtained with up to 0.15 (adipose), 0.20 (brain), 1.0 (heart), 0.6 (kidney), and 0.3 (liver) mg of protein. After 20 min of incubation at 37", snake venom 5'-nucleotidase (50 pg of protein, Sigma) was added, and the incubation was continued for another 10 min. The reaction was ended by 50 ~1 of 55% trichloroacetic acid, and aliquots of the deproteinized incubation mixtures were analyzed for Pi as described previously (16). Butyryl CAMP Deacylation-The assay of butyryl CAMP for all deacylase activities was 37". Reactions were ended by deacylation is based upon the quantification of the [3H]butyric acidification to pH 2 with 2 N HCl followed by addition of 100 acid produced by enzymatic hydrolysis of butyryl-labeled 13H]-pmoles of carrier butyric acid. Incubation mixtures were Btz-CAMP, N6-[3H]Bt-cAMP, and 02'-[3H]Bt-cAMP. In a rapidly extracted three times with 2-ml portions of diethyl ether. total volume of 0.5 ml of 40 rnbr Tris-HCI buffer (pH 7.5), incu- The pooled etheric extracts were treated with 0.1 ml of 1.2 N bation mixtures contained 0.88 pmoles of MgSO+ excess (0.4 to KOH, and the ether was evaporated by a stream of air or nitro-0.7 pmoles) substrate, and various amounts of particulate or gen. Following acidification of the residue at 0' with 10 ~1 of 2 cytosol protein.
Generally, zero order kinetics with respect to N HzS04, t3H]butyric acid was extracted into a solution of 1% either substrate were obtained with up to 0.5 (adipose and liver), I-butanol in n-hexane and isolated by chromatography on col-0.7 (brain and kidney), and 2.0 (heart) mg of cytosol protein.
umns of silicic acid-bromcresol green, as described previously Incubations were usually for 30 min since linearity of deacylation (12). Aliquots (5 ml) of the eluate fractions were counted by occurred for at least this period for all preparations.
The pH liquid scintillation spectrometry in Nuclear-Chicago equipment. optima for N6-butyryl amidohydrolase activity were 7.5 (adipose, Appropriate controls, i.e. incubation mixtures without enzyme heart, and liver) and 8.5 (brain and kidney).
Optimal pH carried through the entire procedure, were performed in order to values for 02'-butyryl esterase activities could not be determined assess the nonenzymatic formation of [3H]butyric acid during with accuracy because of the lability of this ester bond under the incubation and isolation procedures. This minor correction alkaline conditions; however, pH 7.5 appeared to be optimal in amounted to, at most (with [3H]Bt2-cAMP), less than 4$$ of all cases relative to lower values. The optimum temperature that produced enzymatically. the tissues studied are targets of hormones whose actions appear to be mediated by cyclic nucleotides.
Although 02'-[3H]Bt-cAMP was not available for these early studies, later work with this compound revealed that the subcellular distribution of 02'butyryl esterase activity resembled that observed with the dibutyryl derivative.
In all tissues, it was clear that W-butyryl amidohydrolases were essentially cytosol enzymes. In contrast, although O*'-butyryl esterase activities (inferred in Table  I by comparing [3H]Btz-cAMP with N6-[3H]Bt-cAMP) were chiefly soluble in these tissues, substantial activity also resided in well washed kidney and liver particulate fractions. The subcellular distribution of CAMP phosphodiesterase was also determined in the same tissues. Cytosol phosphodiesterase activities were substantially greater than particulate values in brain, heart, and kidney, but phosphodiesterase activities were about equal in both subcellular fractions of adipose tissue and liver (Table I).
Since vasopressin-sensitive adenylate cyclase is located mainly in kidney medullary tubules, whereas parathoromone-sensitive CAMP synthesis is primarily a cortical phenomenon (19), it was of interest to compare deacylase and phosphodiesterase activities in both anatomic compartments; such an experiment is shown in Table I. As the data reveal, soluble W-butyryl amidohydrolase and 02'-butyryl cstcrase activities appear to be equally distributed between the cortical and medullary regions.
In contrast to this symmetric distribution, the specific soluble CAMP phosphodiesterase activity of the medulla was three times that of the cortical enzyme; however, even the cortical enzyme had a specific activity second only to the brain cytosol enzyme. Therefore, intrarenal barriers to the deacylation of butyryl derivatives of CAMP do not appear to exist in the rat.  Particulate and cytosol fractions were prepared as described under "Experimental Procedure." For both assays, incubations were at 37" for 30 min. Phosphodiesterase and deacylase activities were assayed also in cytosols of kidney cortex and medulla; linear phosphodiesterase activities were obtained with up to 0.20 and 0.30 mg of protein of medullary and cortical tissue, respectively; double these amounts of cytosol protein were used in the deacylase assays. groups hydrolyzed per mg of cytosol protein in 30 min. In contrast, phosphodiesterase activities were more variable among tissues, varying over an order of magnitude.
The major result to emerge from these assays, howevrr, is thr marked contrast between rates of phosphodiesterase activities and those of deacylation.
Deacylase activities relative to phosphodiesterase activities ranged from a high of 1: 14 (hcart, [3H]Bt~-cXMP) to a low of 1: 185 (brain, NG-~3H]l~t-cAMP). DISCUSSION Our preliminary (10) and present studies demonstrate directly that cell-free extracts of rat adipose, brain, heart, kidney, and liver contain both O*'-butyryl esterases and NG-buytryl amidohydrolases which, when acting sequentially or in concert, can hydrolyze Btz-cAMP to either of its monobutyryl analogues or ultimately to CAMP. In general, this deacylation was essentially a cytosol phenomenon, although liver and adipose tissue particulate fractions also exhibited substantial O*'-butyryl esterase activity.
From the point of view of the catabolism of acyl derivatives of cAMP to inactive products, e.g. 5'-AMP, it is significant that CAMP phosphodiestcrasc activity was also highest in the cytosol fraction of these tissues. Thus, no anatomic barriers appear to be present to the cat.abolinm of these derivatives, i.e. deacylation and opening of the cyclic phosphate ring. Of ancillary interest is our obstrvntion that, in adipose tissue and liver, the cytosol and particulate phosphodiesterase activities were similar. Substantial (60"1, of the total) particulate phosphodiesterase activity has also been noted in beef heart homogenates by Sutherland and Rall (20) and in brain by DeRobertis et al. (21) ; however, io both instances: although the soluble enzyme had the higher specific activity, it did not appear to otherwise differ from the particulate enzyme. Since butyryl derivatives are ineffective as substrates for CAMP phosphodiesterase, it is clear that deacylation of these derivatives is a prerequisite for further catabolism by phosphodiesterase.
Since (Table II) deacylase activities in all tissues examined were lower than phosp~~odiesterase activities in the same tissue compartments by from one to two orders of magnitude, and since 5'-nucleotidase activities in whole homogenates of at least one of these tissues, namely, liver, averages about 2000 nmoles of 5'-AIV'IP hgdrolyzcd per mg of protein in 30 min (100.fold greater than deacylase activitie?), it is also clear that deacylation is the rate-limiting process in the catabolism of di-and monobutyryl derivatives of cXR4P to adenosinc.  In studies with fetal rat calvaria, Heersche ef al. (13) observed that Btz-cAhII1 (initial concentration, 0.3 to 0.6 rnM; intracellular concemrstion after 15 min of incubation with intact tissue, 0.12 nnvr) elevated tissue levels of CAMP 2-to 3-fold during 15 min of incubation from basal levels of about 0.3 nmole per g (wet weight) of tissue. Since deacylation of Bt2-cA4MP under their conditions produced at most 0.2 nmole of CAMP per g of calvaria (13), and since under certain conditions BtQ-cX1LIP (0.1 mM, a 166fold molar excess over the substrate CAMP) significantly (about 30%) inhibited the cAM1' phosphodiesterase activity of calvarian extracts (13), Hecrsche and co-workers concluded that Bt-CAMP acts primarily as an inhibitor (competitive?) of CAMP phosphodicsterase, thus permitting the accumulation of endogenous c4hIP.
It should be noted that, at high, nonphysiological concentrations of CAMP, others (8)  have not yet been eliminated.
(a) Bt-CAMP may act as an intracellular substitute for CAMP. This possibility is supported by evidence (12, 13) that B&CAMP functioned in intact cells without detectable enzymatic deacylation, but is refuted by reports (11,22) that in cell-free, skeletal muscle glycogen phosphorylase and synthetase systems B&AMP, at low, equimolar concentrations, wets only a poor substitute for CAMP.
Of course, this latter observation might also be explained by an absence of deacylase activit,ies from the preparations employed (&!a in&~).
(b) Diand monobutyryl derivatives of CAMP might be deacylated to the active species, namely, CAMP itself or intermediate monbutyryl derivatives.
Our present experiments make it clear that at least five major organs have the potential of producing from di-and monoacyl derivatives of CAMP amounts of CAMP that exceed those produced endogenously following maximal hormonal stimulation of adenylate cyclase. In general, basal concentrations of cAh!Il' in various animal tissues range between 0.1 and 0.5 nmoles per g (wet weight) of tissue, or about 0.5 to 2.5 pm&s per mg of protein (1); hormones increase these values from 3-to IO-fold, depending upon the tissue and hormone (23), and these concentrations are more than sufficient to produce maximal physiological responses (1). In the present exprrimeuts, t,issue dcacylases were capable of producing at least 10,000 pmoles of CAMP per mg of protein in 30 min from butyryl derivatives of CAMP. Clearly, even slight intracellular dcacylase activities (perhaps beyond the sensitivity of methods used for detection) would produce amounts of CAMP sufhcieut to elicit all physiological effects. It would be of interest to det,ermine the factors which elicit the activities of .!VG-butyryl amidohydrolase and 02'-butyryl esterase.
Undoubtedly, intracellular CAMP from whatever source, including deacylation of butyryl derivatives of c,4XIl', is subject either to degradation by CAMP phosphoditsterases or removal from t.he cell. The extent of degradat.iou is controlled by several factors. For rxamplc, it is not yet clear to what extent intracellular chi\Il' is available to its phosphodiesterases.
It has long been recogiiizcd (24) that much cAXI1' i. s sequestered in intracellular compartments; for rsamplc, 60y0 of rat liver chhll' was found to be associated with particles (25). Furthermore, it has been demoilstrated recciitly that the fraction of skeletal muscle (26) and kidney (27) cAhI1' bound to a cgtoplasmic protein (the regulatory subunit of cAhIl'-activated protein kinuse) was not su~teptible to the actioii of phosphodiesterase.
The stability of the rAS\ll'-binding protein complex has been emphasized (26-28), and the rate of dissociation of this complex was shown to be the limiting reaction in the hydrolysis of c,4MP by phosphodiesterase. The recent report by Chambaut et al. (29) of the presence in rat liver extracts of a cAMP binding protein devoid of kinase activity underscores the possibility that one role of the binding protein is t.o regulate the availability of free CAMP in the cell. The intracellular concentration of cAXI1' may also be protected by physiological inhibitors of the cA?(!IP phosphodiesterases. These enzymes are known to be competitively inhibited by the only other known cyclic nucleotide, that is, cyclic GMP (30-33), and there is a recent report of a nondialyzable, heat-stable inhibitor of cXMP phosphodiesterase in oxynitic cells of frog gastric mucosa (34).
With regard to possibility (a) above, namely, that N6-Bt-CAMP, formed from Btz-cAhIP by NG-butyryl amidohydrolase, is the true imitator of c.4MP, two lines of evidence support this postulate.
(a) Posternak et al. (11) found that NC-Bt-CAMP was second only to CAMP as an activator of glycogen phosphorylase in cell-free extracts of dog liver, whereas B&cAMP and 02'-Bt-cAMP exhibited only minor amounts of activity; and (b) Kaukel and Hilz (7) recently reported that N@-Rt-CAMP, but not B&AMP, had a high affinity for a CAMP binding protein kinase from HeLa cells.
Finally, the identity of the two deacylases should be considered. Since no AT"-acyl derivatives of adenine nucleosides or nucleotides are known in a nature, the NG-butyryl amidohydrolase activities observed in the present experiments may be nonspecific. The only ;V-acyl amidohydrolases which have been identified in animal tissues are the general (EC 3.5.1.14) and specific  N-acyl amino acid amidohydrolases, and the AT-fatty acyl sphingosine amidohydrolase (24). Since the latter is a brain particulate (800 X g sediment) enzyme with an acid (4.8) pH optimum, it is unlikely that this enzyme is responsible for the Are-butyryl amidohydrolase activity observed. A likely candidate, however, is the general N-acyl amino acid amidohydrolase described in kidney by Birnbaum (25), which may be a cytosol enzyme (it appeared in the 4,000 rpm suprrnatant fluid of homogenates) and which had an optimum pH of about 7. Naturally occurring 2'-and 3'.fatty acyl esters of the ribose and deoxyribose moieties of nucleotides have not been found.3 It is likely, therefore, that in the present experiments deacylation of O*'-Bt-cA1IP was carried out by one or more nonspecific esterases. hlost animal tissues contain complex mixtures of esterases with overlapping substrate specificities (26), such as the broad specificity carboxylic ester hydrolase (EC 3.1.1.1) or the more specific acetic ester acetyl hydrolase (EC 3.1.1.6) ; this overlap may be due to an indiscriminate bond-breaking mechanism which will act on any ester that can approach the enzyme active center closely.